1B Parameter LLMs — Page 104

math_acc_1.5B

llama-1b-sft-anthropic

ds-svd-muon-adam-1e-6-global_step_80

ds-svd-muon-adam-1e-6-global_step_100

ds-svd-muon-adam-1e-6-global_step_140

ds-svd-muon-adam-1e-6-global_step_200

ds-adam-1e-6-global_step_40

ds-adam-1e-6-global_step_80

ds-adam-1e-6-global_step_180

pdalma_ctx4_dm1_ce01_pr05_ptll32-1b_s2_ckpt_1_of_10_it4

pdalma_ctx4_dm1_ce0_pr05_ptll32-1b_s2_ckpt_6_of_10_it62

pdalma_ctx4_dm1_ce0_pr0_ptll32-1b_s2_ckpt_1_of_10_it4

pdalma_ctx4_dm1_ce01_pr0_ptll32-1b_s2_ckpt_3_of_10_it12

rlvr_llama1_warmstart_bleu_alma_rbz_256_ckpt_7_of_10

sft_qwen15_code200_lr_5e-6_constant_bsz_64_ckpt_1_of_5

pdcd200_cptq15_ce01_pr05_ptq25-15b_omi_c100k_200tok_s8_ckpt_3_of_10_it45

sft_llama1_alma_lr_1e-5_cosine_bsz_64_ckpt_2_of_5

pdcd200_cptq15_ce01_pr0_ptq25-15b_omi_c100k_200tok_s8_ckpt_6_of_10_it226

hapo_dsr_1b

syn01_sft_fft

sft_qwen15_code200_lr_1e-5_cosine_2_epochs_ckpt_10_of_10

bs3v2_qwen1b5_cnndm

mia-target-model

iampreydata-finetuned-colab-20260308-1137

model_sft_resta

Qwen2.5-1.5B-Instruct-Gensyn-Swarm-mute_robust_mandrill

Llama-3.2-1B-distillation-alpaca-5.0-AlpacaPoison-sauce1-PT

Qwen2.5-1.5B-ViInstruct

Llama-3.2-1B-a100-2

delethink-96k-ckpt150

expert_len_MRL4096_ROLLOUT4_LR1e-6_step50

full_sft_5

qwen2.51.5B-chess-sft-2

binary_cosfmt_MRL4096_ROLLOUT4_LR5e-7_step54

merge_cosfmt_MRL4096_ROLLOUT4_LR5e-7_w0.5_pcb

merge_accfmt_MRL4096_ROLLOUT4_LR5e-7_w0.5_tall_mask_ta

merge_lenfmt_MRL4096_ROLLOUT4_LR5e-7_w0.5_tall_mask_ta

Qwen2.5-1.5B-Instruct-Medical-cpt-sft-v2-dpo-v2

llama32-1b-dynamic-dpo-hh-rollout

ds-adam-1e-6-global_step_200

ds-svd-muon-adam-1e-6-global_step_40

ds-svd-muon-adam-1e-6-global_step_120